DRT technology
DRT technology executive summary
DRT technology replaces the abstraction of the individual client distribution of
Internet connections with the abstraction of centrally arranged transient networks.
The technology introduces the concept of platform-independent, neuronized DRT
application networks.
As mantioned above, the technology provides, filtered, and complementary sets of
application networks (managed and free) and the construction and management of
the resulting networks..
The technology introduces and supports the concept of instant operations, content
sending and communication
Technology introduces the concept of application socialization enabling the
automatic or controlled distribution of knowledge representations. This is
accompleshed with distributed implementation and MI-based supervised or
unsupervised teaching.
The technology introduces the concept of an intelligent “web-assistant” or
“personal-webpage” whereby network applications / clients / subnets provide each
other with automated or managed services through the layers provided by the
technology.
Technology can alleviate data protection concerns, economic and social governance
benefits arising from the monopoly position of large technology companies
The application of technology in IoT devices allows the support of industrial
processes with DRT application networks.
The application of technology at the operating system level allows the possibility of
decentralized data storage and retrieval
For cloud-based applications of the technology, standardization and deployment will
provide an advantage for its first implementers.
With the widespread adoption of technology, the prospects for the use of the
Internet, which are not yet known today, may open up.
Imagine the amazing dynamism and efficiency of the fact that different applications running
on computers, phones, and IoT devices can form temporary mixed networks with different
applications running on other computers, phones, IoT devices for a longer or shorter period
of time. Then these networks break down, new ones are born, intertwined, separated,
while the members of the network provide services to each other and other networks
without even touching it by a human hand. Maybe it’s time to lay the technological
foundation for this.
Presentation configuration:
Raspberry PI IOT
Blue network
Green network
WPF app
Yellow network
Foreign server
DRT own server
Prototype configuration:
WPF hello world App
ASP net hello world App
ASP net real App
WPF real App
WPF real App
Cloud service
DRT own server
Foreign server
mail: bb@gate575.hu
DRT technology overview
We would like to bring to your attention a new technology that will hopefully fundamentally change
the way you use the Internet today. We think that everything we know about the Internet today, will
look obsolete if this new technology becomes widespread. New evolutionary processes will launch in
information technology and applications and in everyday life that may not even be apparent.
This guide is a summary of the DRT technology demonstration
Since a technology presentation says ususally basic operational technological details so naturally dry
and sometimes boring. Thus we created an interactiv presentation to help you focus on the potential
of technology. In this interactive presentation, we only touch on the technical details as much as is
absolutely necessary for undestand. Therefore the first part of the presentation will show you how to
use this technology throught interactive presentation.
The cornerstone of the applicability of any technology is to make it as accessible as possible on any
platform. We currently have two prototypes running on two platforms. The second part introduces
these prototypes.
For the sake of completeness, we will start with some facts we all know so that the basic concept will
become understandable later. The simplified using of Internet is as follows: there is a client machine
somewhere (client machine) and in front of it there is an operator and uses an Internet address to
download a web page to a browser running on your machine. Of course, there is a server somewhere
where that web page is stored. On the web page that appears in the browser, the operator can then
perform some interaction. Data may be recorded on the server and the web page (or a part of it) will
travel several times between the server and the browser of the client machine. When client called A
reaches the web page, of course, it is possible that another client, for example B, will navigate there
and also download the same content to
his browser. Up to dozens of clients and
applications (browsers) can read the
same page at the same time. Taken an
example. The illustrated clients (who are
three) downloaded the same content in
their browsers as described above.
(Green wepages) You can see that all
functions work on all three web pages.
This is a demo of a production report
where you can view partners and jobs In
this case, tha basic question of DRT
technology is to what could sets- consisting of
clients (physical machines), applications and
operators- be considered in this abstract way. Could it be, for example, a ’resource package’ where
shared content can be shared, or could it be a „discrete distribution” in which individual operations
can be initiated? Not to exclude the possibilities of the two the essence of DRT technology is to
consider this set of applications and operators as a Centrally Arranged Transient Network.
mail: bb@gate575.hu
It is central because in the cognitive or morphological sense they read the same server content i.e.
the content / server is in focus. Transient because in browsers running on clients, operators can
navigate to or from at any time and then the set is no longer with that number of elements, namely
the set is temporary. Let us also add that no registration is required for this network membership, so
the technical definition is more precisely: this set is a Central Arrangement Transient Network
without Registration. The basic question now is, what would be the benefits of having a technology
capable of building this network? Before we answering this question, let’s refine the model.
Although each client appears to be equal in this network, there is or may be a client whose status is
different from the others, for example, because he is the owner of the website. This status can be
associated with a special right ofl,
for the network this application /
ie
operator
/ client can be the
Own client on the network
network administrator. The former
question can be reworded as
follows: what are the benefits of
being part of a network and / or
supervising it when
I visit a
website? The accomplishment of
such a concept can only be realized
The other members of the network
with
a
single
technology
implemented on all platforms. This
is the DRT technology in which DRT
stands for Dendrites. Let’s look at a
working interactive example. Now, in web applications above, we start building the network on the
three browsers. We see that, perhaps for the first time in the history of the Internet, an application
sees a transient network organized around it self.
Unique name
For later understanding, we briefly describe the elements of visualizing a presentation. Each client is
characterized by a set of properties enclosed in a frame. Hereinafter, the term client, unless
specifically mentioned, always includes the machine, the application and the operator. This will be
important for the extension of technology and the example of artificial intelligence / machine
learning. Each client sees itself in the top right position in the network. A client in this symbol set is
characterized by the following. The first icon in the client header is a unique name. It is also involved
in identifying the internet address as well as a mac number and an internal tag. What we are seeing
now is a network built around the same website but of course it is possible to create a network of
other websites. Therefore, each network also needs its own identification, which is symbolized by a
lamp in the upper left corner of the icon group. We are now part of the green network. There is a
mail: bb@gate575.hu
connection shown on the surface of the lamp. Because of the transient nature of the network, it is
important to continuously display the structure of the network and the presence of members. For
this reason, ther is an existence counter placed in a red frame on a blue background. This is an
incrementally down-counting register that counts down to zero when the presence is terminated and
when the network membership is terminated. (in parallel, we show an example of network
membership termination …)
Since later, by expanding the technology, we build the network in any machine and platform
environment, so it is useful to know what hardware serves the network. Therefore, the current
hardware device is also displayed. We currently support the display of Desktop, Notebook, Tablet,
Phone and IOT devices.
Desktop hardware
Each hardware device can be selected from a combo, below
the symbol set represents the client in the representation. We
also mention that in the presentation we illustrate the network
membership presence as well as the networking
communication with the server
through a continuous
dialogue of randomly trained literals. The Client sends a
question and the server
„Conversation literals”
Server status info
answers something that is
displayed on the client interface while the server provides
information about its own operation.Network membership, as
previously indicated, is currently arranged at two authorization
levels. It can be a client supervisor or a client. The default
membership right is the client, which we do not even display. If a
Selector
Dino eye
Supervisor
Selected client
membership can also have supervisory
authority, it is indicated by a red letter
S in the upper right corner of the client
interface. When using the service layers to be detailed later, it becomes necessary to select a client
for different purposes. This can be initiated by clicking on the small check mark icon on a green
background. The current client interface will then be selected. Again, raising the question of the
mail: bb@gate575.hu
benefits of building, membershipping and overseeing such a network, we have come to the approach
that the potential benefits and service-based functions should be grouped together. We basically
support the management of six service layers of which abbreviations are as follows:
Content Management Layer
Behavior Management Layer
Input Management Layer
CML
BML
IML
RemoteProcess Management Layer
Interpersonal Communication Management Layer
Network Topology management Layer
RML
IPM
NML
(It is important to note that the use of DRT technology does not affect the original functionality of
the host application in any way. DRT technology can be turned on and off at any time. In the
interactive presentation, the solution supports the construction of a DRT network of up to ten
members together with the local client.The prototypes
allow the possibility to create a network of a thousand
member with the help of the so-called „Client Ring
Selector”)
Supervisor mode
You can view each service layer in groups by clicking
on the "dino eye" next to the local client collector
icon.. Of these, the SN-labeled switch and the single
network / multi network switch as a service ensure that the
current green network can connect to other networks..
Single Network Mode
We begin the presentation of each service layer with the Network Topology Management layer. This
allows you to quickly understand how layers work and it will be needed later when we introduce
some DRT-specific concepts. It has been mentioned before that each network has a unique network
ID. (DRT network is still the same webpage now, but later as we introduce the concept of DRT
Application Networks we will see they can consist of any mixed applications). Currently, we are the
members of the green network which is symbolized by the green lights on the clients. This network
(the green network) currently has three members. At
the same time, a yellow network (as well as the azur
network in the presentation) was built around a
completely independent website, which also consists of
three members. Only the supervisor authorization
operator can be allowed to use the
MultiNetWork
Network Topology Layer. If you switch to
MultiNetwork mode in the NML layer of
the green network,
Two networks, green and yellow…,yellow are not yet connected to
Connected to the yellow network
Here you can choose which other network you want to be a member of. If you switch to the yellow
lamp, you can set it as a two-position switch to join or not. If we add, the union of the two networks
mail: bb@gate575.hu
will appear in the green webpages network , which can be seen on the network membership lamps
of each client.The new network (formerly green) thus consists of six members, three green members
and three yellow members. Of course, within the yellow network, you still only see your own
members because we didn’t change network membership there.
Green network has three endpoints
Yellow network has three endpoints
A simple union of two networks (green and yellow) has six endpoints
mail: bb@gate575.hu
So far we have only presented in web environment. Before we describe the service layers of the
present technology, we extend it to other environments as well. We have two intranet desktop (eg
WPF) applications below.
WPF Desktop app (1)
DRT Ready switch
WPF Desktop app (2)
This WPF application has a completely different purpose in terms of basic functionality than the web
application seen earlier. However, thanks to the support of DRT technology implemented on that
own platform, it is able to connect to the DRT network as a transient endpoint. Thanks to the same
basic design principles, the application here is independent from the DRT network here as well. (This
limitation will later be we exceeded in the artificial intelligence and machine learning prototype
example). Let’s suppose we have a green network and let’s say we have two desktop applications on
two other (different)
machines. If we now
launch the DRT network
architecture service in
our desktop applications,
we will find that the built
transient network can
see
the
desktop
applications if they had
been given the same
network identification.
(Desktop
applications
should also be part of
the green network). We
see that the previously
First dektop-application
three-member green network has now become four members and the new member is the first
desktop application. Then, with the launch of the second destop application, the green network now
has five members. A platform independent DRT mixed network was established.
mail: bb@gate575.hu
The next platform on which DRT technology can build a network is the platform of IoT devices. In this
simple IOT application, the flashing
frequency of a series of three LEDs can be
adjusted. As you can see in the pictures,
IOT application
the DRT network recognizes the device and
the device also recognizes the network.
That is, the mixed applications built by DRT
technology can be seen on three platforms.
IOT device Raspberry PI
We introduce the concept of
a
platform-
independent DRT application network here, which
in this sense means that DRT technology is able to
create the network among any applications as
described above. As we will see later, it is possible to
support several types of services. The possibility of
IOT device Raspberry PI
building networks opens up space for joint resource and operational sharing transactions of
applications / clients / operators. We say that networks are –„neuronized”, that is, they are able to
connect to other networks as an operating unit and use their combined resources to serve the
emerging information-, transactional- and operational needs. (I can't find an English term for this. In
Hungarian, it's „neuronizált” which means it works like a neuron in the brain.) This idea is
fundamentally different from the concept of a unique client distribution typical of today.
DRT technology replaces the abstraction of the individual client distribution of
Internet connections with the abstraction of centrally arranged transient
networks.
The technology introduces the concept of platform-independent, neuronized DRT
application networks.
As mantioned above, the technology provides, filtered, and complementary sets of
application networks (managed and free) and the construction and management
of the resulting networks.
mail: bb@gate575.hu
Surely the question arises in many people, and it is important to see now what the practical
consequences may be of building such networks. Take, for example, an illustrative example that at
first might not really seem a lifelike event. For example, supposing, that several user visit four of the
websites of Budapest cinemas at the same time, as shown in the figure below. The M1 cinema
website has been downloaded by three people and they form the red DRT homogeneous (i.e. built
around the same application) network. The endpoints are characterized by two notations and this is
formed from the movie theatre ID and the client ID in the figure. For example, the notation M1C1
means that the client C1 downloading the web page of the cinema M1 includes the operator, the
device and the application (in this case a
web page). Similarly, there are three other
homogeneous DRT networks (green to blue
and brown) as follows. Members of each
homogeneous DRT network see each other
within their own network and can connect
with each other within that network, e.g.,
Green DRT network 2 endpoints
exchange tickets or discuss movie offerings
or use other services. However, the more
significant advantage occurs when the M1C4
client, which is the only member of the
brown network in its own DRT service,
turns on the visibility of other networks.
This client is building a new network with
Red DRT network 3 endpoints
Braun DRT network 1 endpoint
Blue DRT network 2 endpoints
DRT technology. From then on, the
M4C4 client will not only be a member
of the network with the M4 cinema,
but will also be a member of a new
network of nine endpoints, namely
the network of applications, devices
and operators of the other three
cinemas. Moreover, the network
building service allows us to generate
more and more refined subnets with
“plane intersections” of the set (described in more detail) according
to different aspects. If the connection endpoints are considered as
points of a plane with a geometric illustration, we get the attached
figure.
mail: bb@gate575.hu
On this plane, on the upper horizontal line (axis), we have now recorded the web applications of the
cinemas while in the vertical direction (axis) the clients are marked.The plane itself thus represents a
set of cinema web applications and the clients (operators, devices) that download them, which could
be called the "cinemas web plane". If we now add the planes of other similarly bound applications
and clients in parallel with this plane, then the resulting spatial object is a cube of wich horizontal
plane is a “web plane” like the previous ones. For example, let some other entities be organized
around the arrangement of restaurants, theaters, confectioneries, chocolate shops , and so on.
Chocolate shops
All entities’ plane section,
reducated to one client
Restaurants
Theaters
eatres
A subnet created with all
the clients of an entity
Clients
Entities
Each grid point in this cube represents a single entity and client binding that can be designated as
part of a network. The three planar sections create a set of specialized or one entity along each
outer edge. According to the example above, a wide variety of combined networks of applications
can be formed, which becomes necessary to achieve a particular task or goal. The next level of
networking / connection is when (as we will see in the prototypes) the above networks can be set up
without operator initiative for a certain period of time if the application is active. The transitional
networks thus created can then be dismantled, new ones born, intertwined, split up, while ad hoc
members provide quasi-automatic services to each other and other networks. It is certainly obvious
that networks of applications operating on this principle can be a source of many new possibilities
that may not even be visible today. Perhaps it is enough to point out that in the 1970s, when the
first documents passed between two universities, no one could have predicted that this would one
day become “youtube” “facebook” “twitter” and the extensive software infrastructure that is now
commonplace part of our lives.
mail: bb@gate575.hu
Content Management layer description CML.
Let’s look at a DRT network that consists of three web and one desktop application endpoints. The
purpose of the Content Management Layer is to enable applications to provide different mime types
of content to operators or applications. That is, DRT technology allows various content to be shared
in a built-in way. Let's see an example of this. As a general rule of thumb, each DRT network service
always applies only to the designated endpoints that can be
Content management
activated with the green-based selector on the clients.
Select, say, two of the three web clients and one of our
desktop clients. Also, since this feature requires supervisor
privileges, click supervisor mode. If we now request a
content service, we get the following interface. In the text
bubble, you can select some pre-made content from the
combo below and the header
can be edited with Input
Supervisor
Management layer. These are
still richtext based text files in the presentation that can be generated
with a text editor, for example. Select the content named personal
message, and then click the button labeled ’Send’. Since we previously
selected three clients, the submission applies to these clients. We
expect the previously edited content to now appear in the interface of
the selected clients. Of course, this only happens if the clients in
question are members of the network, ie they have turned on the DRT technology service.
Selecteds
The content sent
mail: bb@gate575.hu
hecontentthatappears is a service
e Content Service Layer thus provides a
ssibility of the transmission of various
xt-based content (notifications, personal
public interest information, images,
deos, etc.) in a built in way. In case of
prototypes, it is already in html base
and in an online editable form. In other
words, an operator included in the
concept of a client can provide instant
content to clients of his own or
connected networks from any application that supports DRT technology. It reaches everyone or only
the targeted ones, which can of course be enabled or disabled.
RemoteProcess Management Layer RML .
Remote procedure call management allows you to solve a task
Remote process service
using the resources of a neuronized network. The previous
three clients have to stay selected, i.e. the two web
applications and the one desktop application. Clicking on the
menu gives you the next interface. Of the three screens, the
left one contains the information of the sent job, the middle
one shows the received results and the right one shows the
returned messages. The middle screen is actually a result
interface of three panels organized above each other where
we can expect a returned text, image, or graphic type display corresponding to the given task. Their
visibility can be
controlled
independently with
sliders, and check
boxes next to the
sliders can be used
to show or hide the
response results of
each client. Below
the middle screen,
on the left side,
there is a three-
combo and a push-
mail: bb@gate575.hu
button control with the following functions. In the top combo we can
see the clients whose resources we want to use. You can select the
task from the combo below. There
are actually only three task options
to choose from here. Select the
Complex function RPC item. This
task means drawing a multivalued
function (approximately 7000-10000 pixels) with the selected
clients and then returning the
points of the calculated curves per
client to the sender of the request.
Next, the resulting functions are
displayed on the result interface (since
this is a graphic) on the canvas layer.
This function can be parameterized and the third combo is used for
this. Here you can set the parameters AParam, BParam, Uparam,
Wparam FgnParam and StrokeParam. The calculation of the
multivalued function is performed point by point by the selected
client according to the fineness of the division. Set different
parameters and send the task. The received results will be displayed on the middle screen for a short
time and we can also see the sent RPC request and response info.
It can be seen that due to the different parameterization, three different (colored) function images
were obtained. You can view these one by one and use the Literals slider to view the execution code
itself. This, as mentioned in the introduction, is just a presentation. Prototypes already include a code
editor, debugger and compiler, and you can control whether the task runs on multiple threads.
mail: bb@gate575.hu
Another RPC option is sending instant video or image to selected clients. To do this, now select our
desktop client and ask for a remote image. After waiting for the remote desktop client, the following
capture window will appear and an image will be created, which will be sent to the client issuing the
RPC as requested. The use of RPCs is a fundamental technique in the operation of neuronized
networks, as will be seen in several examples in the description of prototypes.
Remote camera imageat
the place of manufacture
Remote camera image
upon arrival at the
issuer of the request
Interpersonal Communication Management Layer IPM for short
This service includes two-way communication among
network clients, which can take place via a traditional,
chat or online video / audio interface.. We do not limit the
possibility of this, ie all clients are equally able to initiate
conversations or receive such an initiative. In the dedicated
interface, each client is color-coded with client names. Clients can
group memberships in their conversations or ignore the
conversation initiative. A conversation from another client is
symbolized by a prominent icon on the client's interface, which is
Chat message received
the first in the client line. Clicking on this will open the
communication panel and the conversation can begin.
mail: bb@gate575.hu
You can initiate a conversation
by clicking on the
service
panel
icon, which allows you to
specify in the list of current
ts who will
e conversation.
Chat panel in conversation
You can update or delete the
conversation and encode the
conversation for security goals
the for a certain amount of
time. This can only be
unlocked by entering
a
previously set key. There is a
lot of talk these days about the
protection of personal data around the world, and a key issue in these debates is access to
conversations in a protected and unprotected environment by unauthorized people. The solutions
offered by large technology companies, given that they are based on central storage, will always
raise suspicions of subsequent unauthorized access to stored text or voice-based information. The
use of DRT technology basically rules out the possibility of this. Conversations are only stored on
clients or on servers of the same ownership as the clients. Thus, only when these client-owned
servers are physically accessed it is possible to retrieve confidential information. So personal data
will remain in the possession of the owners as opposed to the current practice, when it is usually
owned by a third party (technology company). The possibility of eavesdropping-like access in
unprotected media also arises, but this can be significantly reduced by using built-in concealment
procedures.. Compared to the offerings of technology companies that previously required very
significant development costs (Skype, WhatsApp, Viber, Messenger, etc.), DRT technology may
provide greater security and discretion.
The target audience for the technology, as will be
explained in detail in the business model, is very
wide, as it can be used for all existing websites,
even in retrospect. (over a billion (only) web pages
at the moment). To illustrate that the technology is
able to provide quite surprising (perhaps not even
known today) services on the remote clients of the
network and their resources, we present a so-called
thought-based service. This is actually an EEG-based
Neuro Panel
(i.e. with brain waves, without human contact) control initiated at a point in the network, as a result
of which the other clients of the network perform the set (imagined ?) task using DRT technology. Let
the initial network architecture be a mixed network built by three Internet and one intranet client.
mail: bb@gate575.hu
The intranet client activates the Neuro panel in the Interpersonal Management layer wich shows an
image of a grid and a snail, as well as a progressbar group representing brain waves in the upper left
corner of the panel. The control algorithm can be started using the push button, as a result of which
the progressbar group indicates the intensity of the frequencies according to each brainwave range.
Before starting the task, we place an EEG sensing device that detects each brainwave frequency. Let
the task be to move the snail from the left to the right, so that, for
example, we intentionally increase the intensity of the delta brain waves.
Certain combinations of brain waves are characteristic of the (even)
conscious processes that take place in the brain. Thus, after some simple
exercises we have invented, anyone will be able to control the movement
of the snail only with their brain waves. Producing the brainwave
combination expected by the algorithm developed for this, the snail moves a few pixels to the right.
If the algorithm detects a different wave combination than expected, it will deviate vertically.
mail: bb@gate575.hu
The "control idea" performed on a desktop computer using Dendrite technology reaches one,
several or all clients on the network. In other words, with a little practice, it is possible for the snail to
start from left to right in real time on our own intranet client machine and on other internet clients
on the network, as we c
„Thoughtcontrol ”on two different platforms
With this small example we wanted to prove that the technology allows a very wide range of
network services (possibly the creation of new mime types). We want to provide developers with
standard API-based extensions, opening up the possibility of other uses that are not yet known
today. An interesting possibility is the characteristic we found during the development. Combinations
of brain waves, or a special sequence of them, are characteristic and personally unique to certain
pre-learned operations. That is, it is possible that these sequences, due to their uniqueness, are
suitable for use as a kind of “biometric identifier” or “biometric private key”. Since the development
was basically not aimed at this path, we didn’t explore the options in more detail, but it might be
worth the effort.
Gree (DRT) network
GIROInst operations
Yellow (DRT) network
Finally, an outline of a possible solution to today's current demand using DRT technology. In the
example above, one of the clients in the green network (seller / buyer) connects to the yellow
network and the blue (bank) network and then initiates an immediate order in the blue network at
the bank of the client (seller / buyer).. (The initiated transaction is in accordance with the HCT INST rulebook v2.0 according
to the Instant Hungarian Credit Transfer, SCT Inst payment scheme supplemented with Hungarian features). The result of the
transaction is that the immediate transfer request initiated in the green network is instantly fulfilled
mail: bb@gate575.hu
on the account of the client in the yellow network as the beneficiary. More details based on the
figure. Similar to the above, other instant operations can be imagined e.g. buy, sell, auction,
authenticate, file exchange etc.
The technology introduces and supports the concept of instant operations, content
sending and communication
Description of DRT prototypes
In the second part, the prototype implemented on the two platforms is presented. First of all, DRT
technology can be integrated into any application by placing a simple compact control. By default,
this control takes up very little space and you can only see a label labeled DRT
Ready. Clicking on this will open the entire DRT panel where the operation can be
switched on and off manually (and as we will see later, remotely, automatically). It is possible to hide
the DRT technology panel again while the services are running. In this case, the word Ready appears
in the label with a red underline. When creating the Prototypes, we have already strived to ensure
that the visuals are the same and that all implementations produce roughly the same response times.
Two implementations are presented: one is a WPF implementation and the other is a traditional ASP
framework implementation. We are also working on ASPNet Core and ASPNet MVC as well as other
platform implementations.. In
the prototypes, in addition to
the
“traditional”
network
we
network-
and cloud-
application
also
the cloud.
We tried
to show
the use in
DRT panel control in an AspNet Hello World application
mail: bb@gate575.hu
applications, but we also created a “classic” startup (Hello World) application to illustrate that the
technology is really very easy to apply (even with retrofit “installation”). First let’s have a look at the
ASPNet Hello World implementation.
Now, let’s launch an instance of the WPF Hello world app as well. In addition to the two hello world
applications, let's start with a working real ASP and WPF application example..
controlinaWPFHelloWorldapplication
DRT panel control in a real Aspnet application
mail: bb@gate575.hu
A working example of a web should be the web interface of a boarding house, which is designed as
above.
An example of a real WPF application is the production management system of a printing company.
DRT panel control in a real WPF application
The first new prototype feature we introduce is the management of number of network elements. In
the presentation we coluld only display a network of ten clients, although obviously a transient client
with a much larger number of elements can form the network. The control with which we can
navigate among network members, the so-called Client Ring Selector.
The client ring selector:
In the upper right corner is the Client Ring Selector for selecting up to 1000 clients. Its use is based on
the fact that ten clients can be displayed on the panel at the same
time. Three-dimensional rings with 10 clients each are formed in a
“depth” cover perpendicular to the screen, also in 10 planes. The
red polygon represents 10 clients of a Plane, and the sliders can be
used to select a new Plane in
that Plane or perpendicular to
it. In the attached right image,
the zero client ring of the zero
Plane is selected. In the
following image, the elements
of the third client ring, which is
also accessible by the eighth Plane, appear on the central surface
of the DRT panel. Since 10 Plane choices are possible and one
Plane represents 10 clients with 10 elements each, the total
number of network clients is 10 * 10 * 10, thai is 1000 clients.
The range where we are in is shown by the range marker below the selector which is the range from
830 to 839 according to our example. For the sake of the easier identification, the several „depth”
mail: bb@gate575.hu
client Planes are displayed with different colours. (in parallel, presentation for the use of the client
ring selector)
Behavior Management Layer BML for short
In the Services panel, click the second icon in the first row to access the Behavior
Managegement service. By default, the text CM can be read on it which is the default
interpretation of Client Mode.
After activation, you get a
configuration
interface
where you can select the
supervisor mode, the subnet
ID and the public icon of the
client device. Each subnet ID
is represented by a different
color. In the picture we can
see that the network
currently being built consists
of six clients with green IDs.
So here you can choose a
subnet ID e.g. by clicking on the yellow lamp. However,
this option can only be accessed in supervisor mode,
for which we have to click on the button with the big
green client label, which causes the icon label to
change to SM (supervisor mode). So if we click on the
yellow lamp it will mean that our network membership
will change and we will now be
members of the yellow subnet. If
we do so, we will immediately see
the image of another network
membership in the next network
building cycle, we will be members
of a two-member yellow network.
The membership of network is symbolized by the same lamp
as the graphical client profile. Of course, the individual
subnets can be connected, as we have already shown in the
Network Management service, for example, we can see the
union of two networks in the picture above.
The potential of DRT technology, ”socialized applications”, and artificial intelligence.
mail: bb@gate575.hu
Artificial intelligence, which now seems unstoppable, will certainly be a permanent part of every
application, regardless of platform or hardware, in the near future. We want to introduce a new
concept that can bring
almost
development
unpredictable
and
efficiency gains using DRT
technology. This new
concept is the socialization
of applications. The
possibilities are so diverse,
and the process can be so
complex that it is best to
try to illustrate the
possibilities
with
an
example. Suppose that a DRT network is built, the members of which are now as follows.
As you can see in the image above, you now see six clients featuring web applications, desktop
applications, and applications running on a Rasberry PI device. Have each application a module that
uses / requires artificial intelligence and uses its own knowledge base. Be able to expand its previous
knowledge with supervised teaching. This MI module is
available to all members of the network on this interface with
the same visuality. The essence of the use / knowledge is that
the petals of the flower in the picture can be red, green or
blue. The “knowledge” of the module is that it can predict the
name of the flower according to a rule based on the colors of
the leaves (SDCA). The rule is that starting from the top left
letter, we enter the English initials of the current color of the
letters in a right-handed circle. For example, the flower in the
image on the left is named RRG_Colored according to the rule.
By default, each application starts by being able to "predict"
only the names of flowers whose at least two petals are the
same color. This is the „knowledge” that all network
members currently have. For example, if you click on the
"what flower" button, the module will give the correct
answer "RRG_Colored" in each application. Create a flower
with three different colored leaves in any application. To do
this, use the coloring buttons under the petals. If we ask the
name of the flower thus created, the module gives an
answere "How should I know!" in every application
because it does not currently have this knowledge.
However, in each module there is a way to expand current
knowledge with the ability to predict new elements
through supervised teaching. To do this, we can give a new name to the textbox placed in the upper
right corner. By teaching this new name, it will give the right answer to our question. Of course, this
is only for this one client and will be true for this case and only in this one application. For others, this
“knowledge” is still not available.
mail: bb@gate575.hu
However, DRT technology can provide ample opportunities in cases like this to transfer the
knowledge accumulated by a client to any client on the network. Moreover, by socializing
applications without touching / operating a human hand, applications can connect with each other
and increase their local efficiency by having the necessary knowledge accumulated by other
applications.
To do this, for example, select the two web members of the
network as described above. So now we are preparing to make the
increased “knowledge” of supervised teaching implemented in a
desktop application available and usable for a web application
running on a completely different platform. By clicking the send
button, the new knowledge- “acquired” by the desktop client in an
RPC data model- arrives (described in detail earlier) with the help
of DRT technology. The two web members of the network can use
it immediately.That is, web applications must now recognize a
flower whose leaf color order is RGB.
For the sake of example, we now acknowledge both sending and
receiving with a message, but of course the process can take place
completely automatically. Moreover interactions between clients
are possible to intelligently merge potential knowledge and rule
elements on both sides. In other words, it is possible for any device and platform using DRT
technology to share their
experience of artificial
intelligence and machine
learning as “socialized”
entities. In the attached
image, you can see how one
of the web applications
recognizes the three-color
flower in a way that its did
not previously have this
capability
and
no
intervention was done in
this application.
Finally, some options that
may seem futuristic at first,
but are in fact almost at your fingertips with the widespread use of DRT technology. Imagine that
everything and everyone can have a be so-called. personal website. It is actually a web application
that allows a person to business, operations or search center, etc. represents cyberspace. Consider it
as a kind of digital assistant who is always ready to serve and to whom we ourselves or other
socialized digital entities can give different tasks and messages. The basis of the connection, though,
is always DRT technology. It is possible, for example, that in the morning, when we get up, we
instruct our personal website to look for a service within a radius of about 15 km where our car will
be inspected. The exam should not be more expensive than 30eFt. Also, get us two tickets to a movie
of our choice anywhere in the city. We would like you to book a table for two people in a restaurant
mail: bb@gate575.hu
after the cinema but the restaurant should not be, further than 10 km from the cinema Our personal
website visits other socialized applications supported by DRT technology, asks them questions,
analyzes the answers and then notifies us sometime in the morning that the exam has been arranged
for the price of 25 thousand HUF and the date is next Friday, eight o'clock. He then visits the digital
assistants of the cinemas, selects one from whom to buy our tickets, while analyzing the contact
details of the nearby restaurants in parallel, and finally informs us in the afternoon that he has
managed to arrange this as well. He will inform us that unfortunately the distance between the
cinema and the restaurant is not ten but 11.4 km, gives the route plan and the expected average
travel time. For the latter data, he joined the network of one of the route planning applications and
extracted the traffic conditions data. The real novelty of the above processes is that the socialized
application does not have to know the network connections required for the tasks assigned to it at
the moment of issuing the task. On the other hand, with the ability to connect to different transient
DRT networks, we have seen an example of this before, and using artificial intelligence you can
“explore” or use some kind of search DRT networks, find the necessary networks and thus
successfully solve the tasks.
Content Management layer CML
Using this feature
allows you to send
any edited content to
network
members
based on an html file.
That is, an edited
html file (which can contain
other types of content in
addition to text) can be sent.
To activate this service, click
on the icon at the beginning
of the paragraph, which will bring up the Content service editing and sending interface. That is, we
now see three applications, which can be web, desktop, IoT, and so on. applications running on
different platforms and for different purposes. In the production management application, by
clicking on the scan button, we scan a pre-edited content which is a wedding invitation.
Clicking the menu bar icon or the Send icon
on the main control will transfer the
content to the previously selected clients.
As a result, the submitted content is displayed in the production control, application, and test
application.
The window that contains the content, of course, can be removed with the icon in the upper right
corner. There is a way to edit the content directly, but you can also upload a file created by using an
external html editor (eg Word). Edited content can be saved with position and display size data in a
custom format with the extension ".drtcm".
mail: bb@gate575.hu
Remote Shared Task Execution Service RPCML for short
Use of this feature can be activated with the first icon in the second icon bar. This will open
an RPC code editing interface as follows. The next demo has almost no practical benefits. We
didn’t make it because we thought it was a real need, but because it’s great for proving one of
the general and real benefits of RPCs. This is explained in the Analysis of Execution Times section at the
end of the test.
Currently, there are basically two types of RPC types. One of the so-called standard built-in RPC whose
code is protected and cannot be modified, but can be parameterized. The other RPC is that can be
coded and parameterized freely by the user (e.g. in c#). The RPC code editor and activator is the
interface where you can select the RPC tasks to be sent to each client. Here you can also edit the
parameters and start a local code translation or test run. The edited RPC can be saved, deleted and / or
modified. For example, before sending a detailed description, send a task to your connected seven-
member network. Now, as before, the selected clients have to calutate a multivalued function using
their own resources. It is then visualized and the result is sent back to the request owner. The result of
each run can be displayed simultaneously or separately on a transparent interface. Each client writes a
run or error message log about the received task, which is also returned to the owner of the RPC
request, which it can also display. On the receiving side, the technology provides automatic handling
according to the mime type of the returned results. The currently supported mime types are as
follows.{TEXT, HTML, GIF, JPEG, PNG, MP3, CANVAS, WAV, MPEG, DBF, DRTCOMPERROR,
DRTRUNERROR};
mail: bb@gate575.hu
In the upper third of the code editor, there are three combos of which:
With the first we can select the stored RPC where RPCs starting with Std_ are read-only as shown by the
checkbox next to them (Std_Canvas_RPC) The last RPC is a so-called user RPC whose code
parameterization, purpose or returned result can be freely edited, compiled or tested by the code editor
(User_ToString_RPC).
The second collection is a list of selected RPC parameters that the RPC user is free to edit.
The third collection contains a list of assemblies needed to compile the code. Editing is only allowed for
user RPC.
Below the code editor there is window is an icon bar for each edit, start, compile, and maintenance
operation.
What each icon means
The compile start icon that compiles the source file in the code editor. The result of the
translation is displayed on an information interface as follows
On the compile and edit page, you run RPC for testing and debugging as if it were running on a
remote client.
Turns the translation result window on and off.
The selected RPC is sent to the selected remote clients with the configured parameterization.
Implements the result data of RPCs run on remote clients
mail: bb@gate575.hu
In this interface, the result data
obtained for each mime type and
for each remote client can be
displayed as previously described.
The lower control panel shows a
checkbox line that controls the
uniform display and visibility of the
result data of each client. All client
result controllers, however, include
the visibility of text, html, image,
dbf, video, log, and error outputs
that can be controlled by the slider
line above the checkbox line.The
collections of result data for each client should be thought of as being layered on top of each other.
Now we implement the sending of Std_Canvas_RPC so we set the appearance to receive canvas
(image) and expect seven clients. First, we select the seven remote clients, which are actually only six
remote clients, since we perform the calculation with our own application.
mail: bb@gate575.hu
Now, in the editor and activator interface described earlier, we send the RPC to the remote clients.
The result after a short time is as follows for each client:
Together, these give the following
graphical result if you disable the
visibility of text and audio-video
interfaces in each result data panel..
mail: bb@gate575.hu
You can also view the logs of remote clients by making the
LogResult display layer visible. The code is written so that
each run runs according to rndFlg or according to the
received parameters or sets them randomly, so each run
has a different remote client-side result.
cnvResult.Children.Clear();
bool rndFlg = Convert.ToBoolean(args[0]);
Random rnd = new Random(DateTime.Now.Millisecond);
double AParam = Convert.ToDouble(args[1]);
double BParam = Convert.ToDouble(args[2]);
double UParam = Convert.ToDouble(args[3]);
double WParam = Convert.ToDouble(args[4]);
double IParam = Convert.ToDouble(args[5]);
ColorParam = Convert.ToInt32(args[6]);
cnvResult.Width = 700;
cnvResult.Height = 450;
if (rndFlg)
{
AParam = (double)rnd.Next(-50, 50);
BParam = (double)rnd.Next(-50, 50);
UParam = (double)rnd.Next(-10, 10);
WParam = (double)rnd.Next(-10, 10);
ColorParam = (int)rnd.Next(0, 9);
IParam = 250.0;
}
As mentioned earlier, each network
function takes place not only among
clients, but also implicitly among
applications, which raises countless new
possibilities. Previously, each client
implemented
and
visualized
a
multivalued function using its own
resources, and the result, i.e., the seven
images computed by the seven clients,
was displayed on superimposed layers.
The technology has built-in support for
the execution of divisible tasks in which
the clients perform the received task on
several threads in a programmed or
automatized scale. Using the built-in
actor-based execution, we can instruct
clients to visualize any number of
multivalued functions using their own
resources, similar to the previous
example.
mail: bb@gate575.hu
Let the reconstructed network be as follows, where are only two clients form the DRT network as
opposed to the previous six clients.
Selecting both clients for execution and selecting remote execution displays the previous RPC code
editor and activator. Select Std_ActorCanvas_RPC from the list to send the task to the clients. That
is, each of them, one by one, should calculate and draw the seven functions of the example. After
submission, clients perform
each visualization task on
separate
threads,
asynchronously, and each
result is displayed on a
single canvas and then sent
back to the request issuer.
That is, now one client
calculates as many pixels as
the seven clients in the
previous example together,
so 70,000 pixels are
calculated per client, for a
total of 140,000.
The canvas generated by the first performing client contains 70,000 pixels and is shown in the figure
mail: bb@gate575.hu
below. These pixels are seven representations of the multivalued function described in the first
example on a canvas.
The
second-performing
client
returned
with
the
image
below.
mail: bb@gate575.hu
You can see that the images are much more complex than the individual results in the previous
example, as they contain eaxctly seven times as many points.
The following figure shows the running log of one of the clients with the thread IDs of each actor.
The RPC capability in DRT (networked) technology allows for shared task execution across different
device platforms.
Analysis of implementation times:
It can also be seen from the log file that the total time of work done by a thread would have taken
approx. 333 seconds if the client had performed the calculation in series. Thus, it took only 57
seconds to execute on separate threads in an asynchronous manner. The increase in efficiency is
therefore significant. (582%).
Create and run user RPC
The number of user RPC code parameters and their use can be freely edited on the submission page.
As an example, here is an RPC that displays a MessageBox on clients with the code below.
mail: bb@gate575.hu
using System;
using System.Windows;
using DRT_CommonCore.RPCClasses.RPCBases;
namespace DRTCodeTemplate
{
public class RPCTest : RPC_Request
{
public override RPC_Response RPC_Executer(string[] args)
{
string p = args[0];
// Parameter passing
MessageBox.Show("Hello DRT World! param:" + p,"User_Message_RPC process",
RPC_Response retval = new RPC_Response(); // Visszaadott objektum
retval.RPC_RequestName = "User_ToString_RPC"; // Az rpc kérés neve
MessageBoxButton.OK);
retval.RPC_ResponseLog += "\nClientName:" + ClientName + "\n
" +
"request : User_ToString_RPC" +" \n
"
+
"executer running. at:" + DateTime.Now;
retval.RPC_ResponseData = "User_ToString_RPC ok! [" + p + "] Végrehajtva: " + DateTime.Now; //
Visszaadott string
retval.RPC_ResponseMimeType = DRT_CommonCore.Helper.DRTMIMES.TEXT;
return retval;
}
}
}
As you can see, all user RPCs come from an RPC_Request class. By overriding the RPC_Executer
(string [] args) method of this class, we can add our own code content to the RPC to be sent. In this
example, we use only one parameter and display a MessagBox in which the "Hello DRT world" is
written and the resulting parameter is written. Once you have saved your user RPC code, you can still
compile and test it on the sending page.
mail: bb@gate575.hu
In the program we expect a parameter which can be added by clicking on the (+) icon in the middle
collection.
It can be seen that the parameter is “Sent: 26.05.2016. 10:23:00 ” literally. You can then start the test
run with the "gear" icon on the edit page. As a result, the compilation runs on the edit page and
control is applied to the user's RPC code as if it had been initiated by a remote client.
Applications do not have to be the same applications and do not have to run on the same platform,
which means that the technology has a very wide range of applications.
mail: bb@gate575.hu
From now on, user RPC can also be run on clients using DRT technology, as shown on the screen of
our host and test application as we have sent the user RPC to ourselves.
mail: bb@gate575.hu
The application of the technology in IoT devices opens up particularly great prospects. IoT systems
running as socialized applications as described above, creating of the possibility of platform free
intelligent network transactions that are still unimaginable in industrial process control today. We
are also working on such a presentation.
Dendrite is an operating system
The principle of being able to build service-based transient networks and implement arbitrary
operations among their endpoints, and that these network memberships can be freely changed,
further raised the possibility of an operating system that includes the ability to manage / build DRT
networks. A long-standing and logical question is how long and with what efficiency can the storage
infrastructure monitor the amount of information generated? Technology companies today operate
and deploy an amazing amount of servers every day to keep up with the needs. You can still follow
this strategy for a while, but certainly not indefinitely. Perhaps, with another strategy that
automatically expands with the network, it would be possible to resolve this contradiction. For
example, if each newly established network endpoint had built-in storage capacity built into it, the
network would be about to scale its storage capacity to the growing demand. A technological
solution could be, for example, if
the devices outsourced by the
service providers and the user’s
own DRT server contained storage
capacity that a suitable technology
could access and save and retrieve
data from. In other words, large
centralized data storage would be
replaced or much supplemented
by the shared use infrastructure
installed on network service
endpoints, of course, while
ensuring mandatory data security.
One possible solution to this is an
operating system that would support the construction of transient networks, as explained earlier. For
this, we have developed a modest presentation based on a credit card-sized computer, the
Raspberry Pi, which is so fashionable today. Since WinIOT provides the ability for our own application
to “dominate” the hardware, it is expedient to create a small “operating system” on this platform to
present our ideas. Some of these small computers with their operating systems would be able to
serve incoming requests as occasional transient network endpoints.
The application of technology at the operating system level allows the possibility of
decentralized data storage and retrieval
mail: bb@gate575.hu
Dendrite technology and the outside world
The aforementioned services can be extended among the
network members to the immediate environment of each
network
endpoint to
solve any
regulation or
control task
that is within
the scope of
the given client
machine. It is
even simpler
to use special
hardware
according to
the previous
paragraph
supported by
DRT
technology.
In the following, we present a Raspberry Pi 3 application with which we perform control tasks as a
member of the transient network, namely, flashing three LEDs on a test panel, while the device is
also a member of a DRT network.
IOT device on the Web app
mail: bb@gate575.hu
Desktop device on the IoT
The images above show that the small operating system running on the IoT device sees the
applications on the DRT network and treats their services on an equal footing. It is also able to
provide services to the network. e.g changes the flashing frequency of the LEDs controlled by the
device.
Technology introduces the concept of application socialization enabling the
automatic or controlled distribution of knowledge representations. This is
accompleshed with distributed implementation and MI-based supervised or
unsupervised teaching.
The technology introduces the concept of an intelligent “web-assistant” or
“personal-webpage” whereby network applications / clients / subnets provide
each other with automated or managed services through the layers provided by
the technology.
Technology can alleviate data protection concerns, economic and social
governance benefits arising from the monopoly position of large technology
companies
The application of technology in IoT devices allows the support of industrial
processes with DRT application networks.
mail: bb@gate575.hu
Fine-tune the operation of DTR technology:
The tunable functions are based on the following:
1. You can set where the life expectancy calculation for each client should be calculated. It is
possible to configure server-side or client-side lifecycle management.
2. It is possible to manually set the life cycle interval timer from 1000 ms to 100000 ms
3. You can set the above-mentioned “watch dog” interval to be set manually or automatically
depending on the load.
4. The storage strategy of each client's "pings" can be selected to be stored in a database on
the server or in a memory cache in the server's memory.
5. You can set the client's own "ping" interval, ie the time during which an existing client
cyclically registers on the network.
6. It is possible to remotely start or stop the DRT
component used on each client at a specified time. This
means that the registered client automatically starts the
network set-up at a given time and is available to the
network or disconnects from the network at the
appropriate time
7. .
The above settings are only available in Supervisor mode and are
displayed by clicking on the “Tobi” dog icon representing the
“watch dog”. When we order the structure of the network, the
continuous operation of the “watch dog” is indicated by the
animation of the Tobi dog. Each setting function is shown in the
following figure.
mail: bb@gate575.hu
Client cyclic registration
Tobi's WatchDog operation
and tuning -> off / on
WatchDog update setting
Manual or automatic
management of WatchDog
Lifecycle management (server / client)
DRT client wake up
registration
Storage strategy Database
/ Memory cache
Current alarm delay
mail: bb@gate575.hu
DRT technology in the cloud
We have also developed cloud-based support for DRT technology, where the network structure is
implemented by a (possibly) automatically scalable cloud application. Let's see an example of this.
mail: bb@gate575.hu
DRT technology as a new chapter in the evolution of the internet / informatics.
As we wrote in the title above, think of this technology as the next evolutionary stage of the Internet
/ informatics. Initially, there were independent computers whose operating systems supported the
running of one program at a time. They have been replaced by time-sharing operating systems that
have increased resource utilization by running multiple programs at once. This was followed by the
age of local area networks, in which individual computers formed local area networks that allowed
network members to share each other's resources and exchange data with each other in almost
innumerable ways. Another milestone was the advent of the Internet, which allowed clients on local
area networks to operate as members of a vast worldwide network. Dendrite technology, as the
next evolutionary station, creates the possibility for each intranet, desktop and web, IOT_s client
(neuronized) transient application network to be able to organize in a specially cohesive, variable
logical unit once again a level higher than before. The emergence of rapidly evolving artificial
intelligence-based applications makes the capacities described above particularly relevant. Solving a
task using DRT technology can be accomplished in a way that can be automatically scaled without
touching by the human mind and hand by discovering and using ad hoc network resources. In this
case, the application can "find" the "resource networks" needed for the solution, so that the user
does not even perceive it, but only gets the result ready with a short response time.
Imagine the amazing dynamism and efficiency of the fact that different applications running on
computers, phones, and IoT devices can form ad hoc mixed networks with different applications
running on other computers, phones, IoT devices for a shorter longer period of time. Then these
networks break down, new ones are born, intertwined, separated, while the members of the
network provide services to each other and other networks without even touching a human hand.
Maybe it’s time to lay the technological foundation for this.
For cloud-based applications of the technology, standardization and deployment
will provide an advantage for its first implementers.
With the widespread adoption of technology the prospects for the use of the
Internet, which are not yet known today, may open up.
mail: bb@gate575.hu
mail: bb@gate575.hu